摘要：随着Web2.0的到来，用户规模不断加大，信息更迭不断加快，社交网络用户寻求信息变得越发困难，如何帮助用户过滤无效信息，提供符合用户个性化需求的个性化推荐服务是目前学术界和工业界关注的研究重点。针对这些问题，本文整理出相关学者研究出的个性化推荐方法，重点讨论了社交网络的相关理念、协同过滤推荐算法、基于社交网络的协同过滤推荐算法、基于时间感知和用户反馈的个性化推荐方法，通过这三种方法比较研究，发现基于社交网络的推荐算法的效果比传统的推荐算法的结果高，基于时间感知和用户反馈的个性化推荐算法，通过选择合适的参数，可以显著提高推荐方法的质量。

关键词：社交网络；协同过滤；个性化推荐方法

Research on the Present Situation of Personalized Recommendation  of  Social Network

Abstract：With the arrival of Web2.0, users continue to increase the scale, information changes continue to accelerate, social networking users to seek information becomes more difficult, how to help users filter invalid information to meet the inpidual needs of personalized personalized service is the current academic And industry focus research. In view of these problems, this paper summarizes the relevant methods developed by the relevant scholars, and introduces the related concepts of social networks, the collaborative filtering recommendation algorithm, the collaborative filtering recommendation algorithm based on social network, the time-sensing and user's personalized recommendation feedback Using these three methods, it is found that the recommendation algorithm based on social network is more effective than traditional recommendation algorithm. Based on time recommendations and user feedback, we can improve the quality of the recommended method by selecting the appropriate parameters.

Keywords： Social network；Collaborative filtering；Personalized recommendation method

目录

0引言 1

1绪论 2

1.1研究背景和研究意义 2

1.2国内外研究现状 3

2社交网络 4

2.1社交网络定义 4

2.2社交网络特点 4

2.3社交网络理论基础 6

2.4本章小结 8

3协同过滤推荐算法 9

3.1基于用户的协同过滤推荐算法 10

3.2基于项目的协同过滤推荐算法 14

3.3本章小结 15

4基于社交网络的协同过滤推荐算法 16

4.1信任度计算 16

4.2算法介绍 19

4.3本章小结 23

5基于时间感知和用户反馈的个性化推荐方法 24

5.1动态社交网络和用户反馈的定义 24

5.2基于时间感知和用户反馈的推荐方法 26

5.3本章小结 29

6.社交网络个性化推荐方法对比研究总结 30

结论 32

致谢 33

参考文献 34

社交网络个性化推荐方法对比研究